Currently, treatment and prognostic assessment for non-small cell lung cancer (NSCLC) depend primarily on the UICC/AJCC TNM (tumor, node, and metastasis) staging system. Modern 18F-FDG PET/CT plays an important role in the diagnosis, staging, and restaging of patients with NSCLC and provides three-dimensional (3D) metabolic and volumetric image data. However, the TNM staging system, based on surgical resectability of the tumor, has no mechanism to incorporate tumor volumetric data. Recently, we and other groups have found that baseline metabolic tumor burden (MTB) is a prognostic indicator of survival and is better than tumor standardized uptake value (SUV) measurement, in both surgically and non-surgically treated NSCLC patients, after adjusting for TNM stage, tumor SUV, and other clinical prognostic indicators including age, gender, performance status, treatment type, and tumor histology. Measurement of MTB on PET/CT is practical and reproducible with low inter-observer variability-with a mean concordance correlation coefficient greater than 0.94. It takes about 3.6 minutes to measure whole-body MTBs in each surgically treated patient with reference of clinical PET/CT and CT reports based our retrospective study. However, despite these promising findings, there is currently no effective method to incorporate MTB measurement into the TNM staging system or clinical management of NSCLC patients. To that end, we propose to develop a PET/CT-based volumetric prognostic (PVP) staging system that combines MTB measurements and the current TNM staging system for better prognostication in NSCLC. We propose two specific aims: (1) to develop a new PET/CT-based PVP staging system for NSCLC by developing two different competing mathematical models, and (2) to validate the novel PVP staging system with new datasets of patients from different institutions. Our long-term objective is to synergize the prognostic value of the current TNM staging system and volumetric MTB measurement. The public health significance is expected in the following three areas: 1) to improve patient treatment selection in NSCLC by providing more accurate tumor staging and risk- stratification, 2) to assist clinicians and patients by providing more accurate prognostic assessment, and 3) to help better define patient selection criteria in clinical trials. The innovations of the proposal include the following: 1) the proposed new PET/CT-based PVP staging system for more accurate prognostication in NSCLC is innovative; 2) making use of 3D volumetric and metabolic data from PET/CT for NSCLC staging is innovative; and 3) making NSCLC staging more quantitative with the PVP staging system is innovative.